Testing bad captions file parsing.

modules/videos/captions.py

@@ -1,5 +1,4 @@
 #! /usr/bin/env python3
-import pathlib
 from pathlib import Path
 from typing import Generator, Union
 
@@ -7,14 +6,12 @@ import srt
 import webvtt
 
 from wrolpi.common import logger
-from wrolpi.media_path import MediaPath
-from .models import Video
+
+__all__ = ['get_captions']
 
 
 def get_caption_text(caption_path: Union[str, Path]) -> Generator:
-    """
-    Return all text from each caption of a caption file.
-    """
+    """Return all text from each caption of a caption file."""
     if str(caption_path).endswith('vtt'):
         for caption in webvtt.read(caption_path):
             text = str(caption.text).strip()
@@ -38,15 +35,8 @@ def get_unique_caption_lines(caption_path: Union[str, Path]) -> Generator:
                 yield line
 
 
-def get_video_captions(video: Video):
-    """
-    Parse video captions from the video's captions file.
-    """
-    if isinstance(video.caption_path, MediaPath):
-        caption_path = video.caption_path.path
-    else:
-        caption_path = pathlib.Path(video.caption_path)
-
+def get_captions(caption_path: Path):
+    """Parse video captions from the video's captions file."""
     try:
         lines = get_unique_caption_lines(str(caption_path))
         block = '\n'.join(lines)

modules/videos/lib.py

@@ -14,7 +14,7 @@ from wrolpi.dates import from_timestamp, Seconds
 from wrolpi.db import get_db_curs, get_db_session, optional_session
 from wrolpi.media_path import MediaPath
 from wrolpi.vars import PYTEST
-from .captions import get_video_captions
+from .captions import get_captions
 from .common import generate_video_paths, remove_duplicate_video_paths, apply_info_json, import_videos_config, \
     get_video_duration, is_valid_poster, convert_image, generate_video_poster
 from .models import Channel, Video
@@ -188,7 +188,7 @@ def validate_video(video: Video, channel_generate_poster: bool):
         # Video duration was not in the info json, use ffprobe.
         video.duration = get_video_duration(video_path)
     if not video.caption and video.caption_path:
-        video.caption = get_video_captions(video)
+        video.caption = get_captions(video.caption_path.path)
 
     if not video.size or not video.modification_datetime:
         stat = video_path.stat()

modules/videos/test/test_captions.py

@@ -1,3 +1,4 @@
+import shutil
 from pathlib import Path
 
 from modules.videos import captions
@@ -46,11 +47,11 @@ class TestCaption(TestAPI):
         video1 = Video(title='scream', caption_path=str(self.vtt_path1))
         session.add(video1)
         session.commit()
-        video1.caption = captions.get_video_captions(video1)
+        video1.caption = captions.get_captions(video1.caption_path.path)
         video2 = Video(title='bar', caption_path=str(self.vtt_path2))
         session.add(video2)
         session.commit()
-        video2.caption = captions.get_video_captions(video2)
+        video2.caption = captions.get_captions(video2.caption_path.path)
 
         session.commit()
         session.flush()
@@ -91,3 +92,13 @@ class TestCaption(TestAPI):
             # Matches video2.title
             select_textsearch('bar')
             self.assertEqual(curs.fetchall(), [[2, ]])
+
+
+def test_bad_caption(test_directory):
+    """A caption file that cannot be parsed is ignored."""
+    bad_caption_file = PROJECT_DIR / 'test/bad_caption.en.vtt'
+    test_file = test_directory / 'bad_caption.en.vtt'
+    shutil.copy(bad_caption_file, test_file)
+
+    result = captions.get_captions(test_file)
+    assert result is None

requirements.txt

@@ -15,7 +15,7 @@ sanic-testing==0.8.2
 sanic==21.12.1
 selenium==4.1.0
 SQLAlchemy==1.3.22
-srt==3.5.0
+srt==3.5.1
 websockets==10.1
 webvtt-py==0.4.6
 yt-dlp==2022.1.21

test/bad_caption.en.vtt

@@ -0,0 +1,8 @@
+WEBVTT
+Kind: captions
+Language: en
+
+00:00:00.070 --> 00:00:06.380
+this caption is bad
+
+00:05:08.650 --> 00:05:11.26